A P P L I C AT I O N N O T E

ICP-Mass Spectrometry Authors Zoe Grosser, Ph.D. Lee Davidowski, Ph.D. Laura Thompson PerkinElmer, Inc. Shelton, CT 06484 USA

The Determination of Metals in Cosmetics

Introduction Cosmetics of one form or another have been used since the beginning of time. The U.S. Food and Drug Administration (FDA) defines today’s cosmetics as:

“1. articles intended to be rubbed, poured, sprinkled, or sprayed on, introduced into, or otherwise applied to the human body or any part thereof for cleansing, beautifying, promoting attractiveness, or altering the appearance, and 2. articles intended for use as a component of any such articles; except that such term shall not include soap.” 1 The U.S. FDA has not regulated metals in cosmetics except in the case that zirconium is prohibited in aerosol preparations and mercury can only be used (generally as a preservative) when no other alternative is possible. They do have the authority to protect consumers against contamination deemed to be deleterious to health. Proposition 65 in California requires notice when concentrations in a product may cause exposure to hazardous components exceeding a prudent level. A letter from the Attorney General of the State of California discusses the level of concern and calculates that based on documented lipstick use, a concentration of 5 mg/kg would be the level of concern requiring listing.2 The regulations in Canada include statements about heavy metals, and although low concentrations may be difficult to avoid, concentrations above the levels listed in Table 1 (Page 2) require special scrutiny.

Table 1. Canadian metal concentrations of concern in cosmetics.3 Metal

Concentration (mg/kg)

Lead 10

Measurements in this work are performed using ICP-MS. Elements chosen for examination are commonly considered to be toxic, although there are others that might be considered hazardous as well, such as antimony, included on the Canadian list.

Arsenic 3 Cadmium 3

Experimental

Mercury 3

Samples of a variety of commercially available lipsticks (Figures 1 and 2), nail polishes, and skin creams (for evening tone, lightening age spots) were purchased and prepared in duplicate. Nail polishes were painted onto tared weighing paper and allowed to dry before weighing into the digestion vessel (Figures 3 and 4). Microwave digestion was used (Multiwave™ 3000, PerkinElmer®, Shelton, CT USA) to obtain clear solutions. Six mL of nitric and 0.5 mL of hydrofluoric acid (GFS Chemical™, Columbus, OH USA) were added to PTFE vessels with approximately 0.3-0.5 g of sample and the digestion program shown in Table 2 applied. The samples were then transferred and diluted to 50 mL with ASTM® Type I water. The samples were fairly homogeneous and in a form that allowed a representative sample to be easily taken. If the samples were solids, grinding, blending or other procedures might be necessary to ensure a more homogeneous sample to be measured. Preparing replicate samples will allow us to evaluate if our homogeneity assumption is accurate.

Antimony 5 The European Union (EU) has also developed a list of more than 1000 compounds that are banned for use in cosmetic manufacturing. They further limit the amount of these compounds that may be found naturally-occurring in the product. For example, strontium peroxide may be found in rinse-off hair care preparations, but is limited to containing less than 4.5% as strontium in the ready-to-use product.4 Examining the notices of recall in the U.S. for products manufactured in various parts of the world show that a number of items which have been recalled are cosmetics. For example, skin whitening cream from several Caribbean countries was put on an import alert because the level of mercury measured in the product was 8%.5 Another alert was issued for eye colors including Kohl, Kajal, or Surma. In addition to being an unapproved coloring agent, the material was found to contain lead (Pb). There has also been recent concern about lead in lipstick, uncovered through testing.6 The U.S. FDA has not recalled lipstick and states they will do further testing.7 The guidance mentioned for lead relates to the limited guidance available for food, where a letter stating that 0.1 mg/kg of Pb in candy is the maximum tolerable limit for acceptable children’s exposure. This application note will evaluate the levels of several heavy metals in lipstick, nail polish and skin cream. Based on candy guidance in the U.S. and the limits set in Canada, either graphite furnace atomic absorption (GFAA), inductively coupled plasma optical emission spectroscopy (ICP-OES), or inductively coupled plasma mass spectrometry (ICP-MS) would be appropriate, based on the amount of sample taken for digestion and the final volume of solution.

Figure 1. A variety of commerciallyavailable lipsticks were analyzed.

2

Figure 2. Lipstick was weighed prior to microwave digestion.

Table 2. Microwave digestion program. Step Power Ramp Hold Fan 1

750

10:00 10:00 1

2

1200

10:00

3

0

10:00

1

15:00 3

An ELAN® DRC-e ICP-MS (PerkinElmer, Shelton, CT USA) was used for measurement of chromium, cadmium, lead, mercury, arsenic, selenium, and antimony. The instrumental conditions are shown in Table 3 (Page 3). Mercury was also measured using the SMS 100 Mercury Analyzer (PerkinElmer, Shelton, CT USA) which uses a combustion system, requiring no sample preparation and generating no waste. Instrumental conditions are shown in Table 4 (Page 3).

Figure 3. Nail polish was dried on tared paper before digestion.

Figure 4. The nail polish and tared paper were weighed into the digestion vessel.

Table 3. ICP-MS instrumental conditions.

Table 4. Mercury system instrumental conditions.

Nebulizer

Quartz Concentric

Parameter

Spray Chamber

Quartz Cyclonic

Dry

400 °C for 200 s

RF Power

1500 W

Decomposition

800 °C for 350 s

Integration time

1000 ms (per analyte)

Catalyst

600 °C

Replicates 3

Catalyst wait period

60 s

Reaction Gas for arsenic

O2 = 0.6 mL/min

Gold trap

600 °C for 30 s

RPq for arsenic as 91 AsO

0.5

Measurement

100 s

Oxidant gas flow

350 mL/min

Arsenic analysis employed cell technology to remove the chloride interference arising from either natural chlorine content, salt added to a supplement material or hydrochloric acid used in sample preparation, which can degrade the detection limit. This cell allows a gas to react with the sample, and under controlled conditions, the interference is chemically separated from the analyte. Figure 5 shows how the Dynamic Reaction Cell™ (DRC™) operates. Detection limits are improved compared to other approaches and the precision in varying matrices is more consistent. An instrument detection limit for arsenic measured in 1000 mg/L NaCl was shown to be 2.3 ng/L, comparing favorably to detection limits of 0.6-1.8 ng/L, measured in 1% nitric acid solution, using this technique.8

Figure 5. Dynamic reaction cell (DRC) operates with the PerkinElmer ELAN® ICP-MS to react and remove interferences without allowing competing reactions to add other interferences.

Results and Discussion Sample preparation is an important contributor to the results, especially when it influences the consistency and information provided by the results. An article by several researchers, including the U.S. FDA, discussed the need for a small amount of hydrofluoric acid (HF) in the digest to ensure better consistency and recovery of all the metal, including that which might be protected from other acid attack by a silica particle.9 The question always becomes: is it better to report the total metal content or the portion that might be bioavailable? Even if lipstick is consumed and exposed to hydrochloric acid in the digestion process, lead enclosed in silica particles will not be released and is therefore unable to cause harm. The digestion here used HF and therefore reports a total metal value with more confidence, although it may not be all bioavailable. Several performance criteria were evaluated to ensure the method was operating as expected and the sample preparation had been performed correctly. Table 5 (Page 4) shows the results for NIST® reference material Typical Diet (1548a) digested at the same time as the cosmetics, using the same conditions and acids. Although Typical Diet is not a cosmetic matrix, it has some of the same components. The fat content may not be comparable to all of the skin cream or lipstick types, depending on their formulation.

The mass for selenium measurement was chosen to avoid interference from calcium, which may be present in cosmetics. Since this work was completed, the NexION® 300 ICP-MS has been introduced by PerkinElmer, Inc. and performance is expected to be the same or better than the ELAN DRC-e ICP-MS.

3

Table 5. NIST® 1548a Typical Diet certified reference material. Element and Mass

1548a – 1 (mg/kg)

1548a – 2 (mg/kg)

Average (mg/kg)

Certificate Value

% Rec

Cr 52

0.135

0.118

0.127

–

Cd 111

0.036

0.038

0.037

0.035

Hg 202

< DL

< DL

0.005*

Pb 208+207+206

0.045

0.046

0.046

0.044

104

Se 82

0.231

0.251

0.241

0.245

98.4

AsO 91

0.183

0.192

0.188

0.20

93.8

106

Prepared at the same time as the cosmetics, using the same procedure. * Informational value, not certified. Recoveries of the certified values were within ±20%, indicating acceptable performance. Detection limits were estimated from digested blanks run during the analysis. The standard deviation was multiplied by three and adjusted by the dilution of the solid during digestion to give an estimate in the original cosmetic material. Estimated detection limits calculated for the solid cosmetic material are well below the current compliance limits, allowing confidence when measurements are made near the compliance limits.

Table 6. Estimated detection limits in the solid cosmetic. Element and Mass

µg/g

Cr 52

0.044

Cd 111

0.00069

Hg 202

0.0037

Pb 208

0.0096

Se 82

0.082

AsO 91

0.014

The duplicate digestions were run and average values reported in Table 7 for the lipsticks. It is interesting to note that although several have more than 1 µg/g of lead (which caused a controversy when results were reported in the Poison Kiss document10), it is below the regulated limit in Canada. Other elements that may be of concern are also seen in some samples, such as Cr and Se. Table 7. Lipstick results (ug/g). Element Lip-1 Lip-2 Lip-3 Lip-4 Lip-5 Lip-6 Lip-7 Lip-8 Lip-9 Lip-10 Lip-11 Lip-12 and Mass Cr 52 7.41 7.05 93.3 0.226 18.7 0.858 7.94 5.27 0.645 0.520 1.46 1.46 Cd 111 0.0293 0.0179 0.139 0.0353 0.0264 0.00860 0.0398 0.144 0.0288 0.00897 0.0159 0.00595 Hg 202

0.00944

0.00794

< DL

< DL

0.00977

0.0125

0.0107

< DL

0.00707

0.00422

< DL

< DL

Pb 208 1.38

0.545 0.273 0.0407 0.611 1.12 0.591 0.397 0.321 0.943 0.489 0.700

Se 82

< DL

0.211

< DL

< DL

0.329

0.162

< DL

0.708

1.91

105

0.166

0.108

AsO 91 0.828 0.250 0.411 0.481 0.073 0.123 0.209 0.242 0.0571 0.0586 0.0976 0.0969 Table 8 shows the individual results for two sample digestions for Lipstick 3 and standard deviation and relative percent difference (RPD) between the two measured concentrations. The RPD is less than 5%, which shows good agreement between the duplicate digestions.

4

Table 8. Precision and agreement of duplicate samples (µg/g). Element and Mass

Lip-3 A

Lip-3 B

Cr 52

94.4 92.2 93.3 01.52 2.30

Cd 111

0.138 0.140 0.139 0.00129 1.31

Hg 202

< DL

Pb 208

0.271 0.275 0.273 0.00221 1.15

Se 82

< DL

AsO 91

0.421 0.401 0.411 0.0142 4.90

< DL < DL

Average

< DL < DL

SD

RPD

– –

The lipsticks covered a range of colors, from black to pink, red, purple and brown. The results for the elements measured varied quite a bit between the various brands and colors. The black lipstick (Lip-4) was made in China and was one of the lipsticks lowest in lead and the other metals measured. In no case was the Canadian limit or the limit suggested in California for Proposition 65 compliance exceeded. Several of the lipsticks measured in the Poison Kiss publication10 were also purchased for this work. The result obtained here for lead on Lipstick 5 was 0.611 µg/g and in the report on two separate tubes of lipsticks they measured 0.65 and 0.50 µg/g. The measurements in the publication were also performed using ICP-MS, but the digestion used contained nitric and sulfuric acid and no HF. Therefore, the results may be low or may have less precision for duplicate measurements. Another cosmetic with a different kind of matrix is nail polish and a variety of colored nail polishes of different brands were measured for the same elements and the results are shown in Table 9. Colors ranged from blue to pink and red. Again, small amounts of metals are present in most of the polishes and may cause exposure if nails are bitten or chewed and the polish ingested. Table 9. Nail polish results (µg/g).

Table 10. Skin cream results (µg/g) and post-digestion spike recovery. Element Cream-1 and Spike Mass Cream-1 Cream-2 Cream-3 Recovery (%) Cr 52 0.0318 0.0800 0.00867 109 Cd 111 0.00318 0.00122 0.00194 87.6 Hg 202

< DL

< DL

< DL

111

Pb 208

< DL

0.168

0.0180

97.2

Se 82 0.0198 0.168 0.187 – AsO 91

< DL

0.0128

< DL

86.4

Sb 121 0.00187 0.00483 0.0120 92.4 Spike levels in solution: 4 µg/L, except mercury 0.5 µg/L Post-digestion spike recoveries show that the matrix is not unduly influencing the accuracy of the result. Lead was measured on two separate days, at least a week apart, to ensure the results could be replicated (Table 11). The lead results were present at several different concentration levels in the skin cream and showed good agreement over the two days.

Element NP-1 NP-2 NP-3 NP-4 NP-5 NP-6 and Mass

Table 11. Lead results on separate days (µg/g).

Cream-1 Cream-2 Cream-3

Cr 52 6.58 0.999 2.95 0.800 10.9 1.96

Day 1

< DL

0.168

0.0180

Cd 111 0.0364 0.0290 0.0193 0.0160 0.0832 0.0110

Day 2

< DL

0.161

0.0214

Hg 202

< DL

< DL

0.00965 < DL

0.192

0.00826

Pb 208 0.863 0.947 2.12 0.204 6.03 1.82 Se 82

62.2

128

161

61.7

0.0477 0.397

AsO 91

0.147

< DL

0.266

0.0690

1.91

0.510

The third type of cosmetic examined was skin cream. This type of skin cream was intended to lighten the skin to remove age spots or even skin tone. Three types were digested and measured for the elements of interest. Antimony was added as an analyte for this determination to show that it can be incorporated into the multielement run for Canadian compliance. The results are shown in Table 10.

Skin cream has been the subject of a U.S. FDA Import Alert because high mercury content was found in potential import samples. Mercury measurements were confirmed for the skin creams measured in this work using a dedicated mercury system (Table 12). The detection limit is lower than with digestion because dilution is less and low concentrations of mercury were confirmed. Small samples can be used and rapid screening performed because the system requires no sample preparation and generates no waste. Figure 5 shows the autosampler for the mercury system and the sample measured into small boats. Table 12. Mercury results using mercury system.

Conc. (µg/g)

Cream-1

< DL

SD

Cream-2 0.00240

0.00008

Cream-3 0.00824

0.000044

Three replicate analyses averaged Detection limit: 0.02 ng/g in the skin cream

Figure 5. Skin cream samples measured into small boats for analysis. 5

Conclusions

References

Cosmetics contain a variety of components that can be challenging to digest, including fats and silica-based compounds. Digestion using a microwave with the addition of HF to digest siliceous materials may be necessary. Sensitivity to the resulting decisions to be made with the information, such as the need for bioavailability information, may influence the sample preparation acids chosen. Microwave digestion also provides a clean digestion and retains mercury so that a full suite of elements may be easily examined.

1. Food, Drug, and Cosmetic Act, as amended through 2004, Chapter II, Definitions 21USC 321, http://www.fda.gov/ opacom/laws/fdcact/fdcact1.htm.

ICP-MS is a good choice for determination of low concentrations of analytes, allowing evaluation of toxic and potentially toxic components in cosmetics. More sophisticated interference correction, such as DRC, can remove molecular interferences at the mass of interest, allowing the lowest detection limit to be achieved. This application note has demonstrated the successful analysis of three types of cosmetics for toxic elements and shown good results through replicate analyses and spiked samples. Analysis of a reference material helped to ensure that the digestion step and instrumental analysis were under control.

Acknowledgement The authors would like to thank Lindsay Drennan for the confirmatory data on mercury in skin cream measured with the SMS 100 Mercury Analyzer.

2. Letter from Edmund G. Brown, Attorney General, State of California, to two law groups in reference to Proposition 65 Claims Concerning Lead in Lipstick, http://ag.ca.gov/ prop65/pdfs/Lipstick_Letter-a.pdf (accessed 1-24-2011). 3. Draft Guidance on Heavy Metals in Cosmetics, Health Canada, http://www.hc-sc.gc.ca/cps-spc/legislation/consultation/_cosmet/metal-metaux-consult-eng.php (accessed 1-24-2011). 4. Cosmetic Products Regulation, EU Regulation 1223/2009 (accessed 1-24-2011). 5. Import Alerts, http://www.fda.gov/ForIndustry/Color Additives/ComplianceEnforcement/EnforcementNews/ ucm123909.htm (accessed 1-24-2011). 6. Lead in Lipstick http://www.safecosmetics.org/article. php?id=223 (accessed 1-24-2011). 7. FDA Q&A, http://www.fda.gov/Cosmetics/ResourcesFor You/Consumers/CosmeticsQA/default.htm (accessed 1-242011). 8. Ruth Wolf and Ken Neubauer, Determination of Arsenic in Chloride Matrices, PerkinElmer, (2002). 9. N.M. Hepp, W.R. Mindak, and J. Cheng, J. Cosmet. Sci., 60, 405–414 (July/August 2009). 10. A Poison Kiss: The Problem of Lead in Lipstick, The Campaign for Safe Cosmetics, October 2007 http://www. safecosmetics.org/downloads/A%20Poison%20Kiss_report. pdf (accessed 1-24-2011).

PerkinElmer, Inc. 940 Winter Street Waltham, MA 02451 USA P: (800) 762-4000 or (+1) 203-925-4602 www.perkinelmer.com

For a complete listing of our global offices, visit www.perkinelmer.com/ContactUs Copyright ©2011-2012, PerkinElmer, Inc. All rights reserved. PerkinElmer® is a registered trademark of PerkinElmer, Inc. All other trademarks are the property of their respective owners. 009691A_01